Cloth Simulation Filtering Method Combined with Axis-Angle Transformation

Aiming at the problem that existing LiDAR ground point filtering algorithms have poor adaptability to complex terrain and ground objects, a cloth simulation filtering method based on axis-angle transformation is proposed on the basis of the Cloth Simulation Filtering (CSF) algorithm. This method first preprocesses the original point cloud to achieve downsampling and denoising, then segments the point cloud according to slope information; by introducing axis-angle transformation and combining the orthogonal projection formula to horizontalize the target point cloud, the CSF algorithm is adopted for point cloud filtering to realize the accurate extraction of ground point clouds. Experimental results demonstrate that the proposed improved algorithm outperforms the original CSF algorithm in complex terrain areas, with significantly reduced TypeIerror,TypeIIerror, and total error.